I’d like to have a boxplot for each column of my dataframe (example), with pandas it’s as easy as df.plot.box*()*, but with Bokeh I must specify the categorical variable, even if it is an optional parameter.
`from bokeh.charts import BoxPlot, show
import pandas as pd
Here’s the error message: “null is not an object (evaluation ‘h.indexOf’)”
···
On Thursday, May 4, 2017 at 6:21:29 PM UTC+2, Enrico Rotundo wrote:
I’d like to have a boxplot for each column of my dataframe (example), with pandas it’s as easy as df.plot.box*()*, but with Bokeh I must specify the categorical variable, even if it is an optional parameter.
`from bokeh.charts import BoxPlot, show
import pandas as pd
Bokeh is not the same as pandas. They have different nyerfaces, and the labels parameter is not optional, it is required to specify what column to do a groupby of:
If you’d like BoxPlot to behave differently then you can file a feature request issue in the bkcharts repo, which will soon be the new home of bokeh.charts:
Here’s the error message: “null is not an object (evaluation ‘h.indexOf’)”
On Thursday, May 4, 2017 at 6:21:29 PM UTC+2, Enrico Rotundo wrote:
I’d like to have a boxplot for each column of my dataframe (example), with pandas it’s as easy as df.plot.box*()*, but with Bokeh I must specify the categorical variable, even if it is an optional parameter.
`from bokeh.charts import BoxPlot, show
import pandas as pd